Jump to content
You need to play a total of 20 battles to post in this section.
__Riptide_

Does winrate reflect on how a player performs with a ship???

95 comments in this topic

Recommended Posts

23
[NOKAP]
Members
66 posts
6,074 battles

Title,

Here's an example, I have a garbage WR in my harugumo at 47%,, however at 81k average dmg per match, but does that mean that I suck in harugumo? I find it stupid that a player just judge another's stats by simply looking at the WR, They surely realize that not everyone are/ wants to be super unicum that can carry 11 other players so having a below average winrate isn't entirely their fault?

 

haru stats.PNG

  • Boring 1

Share this post


Link to post
Share on other sites
2,708
[WPORT]
Members
7,183 posts
12,195 battles

People who are more well-versed in the subject of statistical analysis have previously pointed out that the win-rate stats and others are basically "bravo-sierra" and not worth stressing about or bragging about.

Edited by Wolfswetpaws
  • Cool 7
  • Meh 3

Share this post


Link to post
Share on other sites
4,043
[WOLFG]
Members
11,188 posts
10,423 battles

Winrate is just one of a number of statistics that can be used to judge performance, and most of the stats can be rigged through how you play.

Either way, 19 games isn't enough to judge any statistics very accurately.

Share this post


Link to post
Share on other sites
89
[ODIN]
Members
216 posts
8,308 battles

Statistics always need to be looked at in context, and together with other statistics in order to get an accurate interpretation. Also, you need to have a statistically significant amount of data points. Hence why only looking at WR or only average damage is a poor way of determining the actual player skill.

Share this post


Link to post
Share on other sites
590
[BOTO]
Members
1,419 posts
17,041 battles
20 minutes ago, __Riptide_ said:

Title,

Here's an example, I have a garbage WR in my harugumo at 47%,, however at 81k average dmg per match, but does that mean that I suck in harugumo? I find it stupid that a player just judge another's stats by simply looking at the WR, They surely realize that not everyone are/ wants to be super unicum that can carry 11 other players so having a below average winrate isn't entirely their fault?

 

haru stats.PNG

No, WR is only part of the picture.

Share this post


Link to post
Share on other sites
Members
1,226 posts
5,930 battles

Over a large number of games if you are good in a ship you will generally win in it so solo win rate is a reasonable measure of skill. That's over a large number of games though. With small numbers of games it's hard for any statistic to really say whether you are good or not.

Share this post


Link to post
Share on other sites
Members
996 posts
11,306 battles

If just one battle had gone the other way you would have a 53% WR in Harugumo, and you probably would not be posting.  Given how close you are to 50% with so few battles it does not mean a whole lot.

  • Cool 4

Share this post


Link to post
Share on other sites
3,118
[-K-]
Beta Testers, In AlfaTesters
7,505 posts
9,777 battles
32 minutes ago, __Riptide_ said:

Title,

Here's an example, I have a garbage WR in my harugumo at 47%,, however at 81k average dmg per match, but does that mean that I suck in harugumo? I find it stupid that a player just judge another's stats by simply looking at the WR, They surely realize that not everyone are/ wants to be super unicum that can carry 11 other players so having a below average winrate isn't entirely their fault?

 

haru stats.PNG

Sample size of 19 is too small to be meaningful.

Read this: https://en.wikipedia.org/wiki/Sample_size_determination

32 minutes ago, __Riptide_ said:

I find it stupid that a player just judge another's stats by simply looking at the WR, They surely realize that not everyone are/ wants to be super unicum that can carry 11 other players so having a below average winrate isn't entirely their fault?

When you say "judge" what exactly do you mean? For me, it's making a judgement on the expected impact that the player will have on the match. Winrate, given enough samples (and keeping in mind other factors, e.g. solo vs div, tiers, etc.) is a very good predictor.

And of course not everyone wants or can be a super unicum. That's beside the point.

Share this post


Link to post
Share on other sites
1,535
[PSA]
Members
5,113 posts
3,732 battles

WR matters when you have a large enough number of matches to account for blowout matches. Unless you're a unicum, individual player contribution is only a small variance for better or worse.

My WR is around 52% across nearly 2900 random matches. My WR in Republique after 10 matches is 30% despite having 100k average damage and a 2.1 K/D.

I abhor stat shamers as a general rule, but people who stat shame on individual ship performance piss me off tremendously as WR in a ship only becomes a good indicator of skill once you've played a good deal of games in it.

  • Cool 1

Share this post


Link to post
Share on other sites
3,118
[-K-]
Beta Testers, In AlfaTesters
7,505 posts
9,777 battles
31 minutes ago, Wolfswetpaws said:

People who are more well-versed in the subject of statistical analysis have previously pointed out the the win-rate stats and others are basically "bravo-sierra" and not worth stressing about or bragging about.

People who claim that win rate is always meaningless as a player performance predictor don't understand statistics and the theory of probability.

  • Cool 6
  • Thanks 1
  • Boring 1
  • Meh 1

Share this post


Link to post
Share on other sites
SuperTest Coordinator, Beta Testers
6,662 posts
12,100 battles

19 games is not enough. In some games, your maximum carry capacity cannot be reached. I'd say at least 50 to 100 games.

Share this post


Link to post
Share on other sites
120
[F4E]
Members
261 posts
5,969 battles

As others have pointed out your sample size is far too small for WR to be a meaningful indication of your performance. 

I would say under 75 games, other stats would be a better indication, such as avg damage, kills, etc.  These stats are generally not as influenced by your team mates.  Still influenced, just not as much.

Over 75 games, WR becomes the much more reliable indication of performance and average damage/kills less so.  WR starts to measure your influence on the win.  There are a ton of examples of amazing gameplay that can really influence a game that aren't measured by other stats.  But they will always affect your influence on winning and losing over a large sample size.

Example: I'm playing CV and lining up a torp run on a BB.  A friendly Shima launches all his torps at the same BB.  As a CV player I can line up on his broadside and get 10K damage, but likely the BB will turn into my torps which will cause him to miss a lot of the Shima torps.  Or I could launch at his bow or stern so he stays on his current path towards doom from the Shima torps.  The broadside torps would net me 10K damage, but not sink the BB.  The Bow/Stern torps would net me 0 damage, but the BB would be sunk.  

Over a large number of games WR incorporates things like, 1) how meaningful is your damage, is it damage against destroyers at the start of the match, or damage farming against BB's at the end of the match.

My estimate of 75 games is just a rough number I use. 

  • Cool 1

Share this post


Link to post
Share on other sites
986
[HC]
[HC]
Beta Testers
3,379 posts
13,062 battles
41 minutes ago, __Riptide_ said:

Title,

Here's an example, I have a garbage WR in my harugumo at 47%,, however at 81k average dmg per match, but does that mean that I suck in harugumo? I find it stupid that a player just judge another's stats by simply looking at the WR, They surely realize that not everyone are/ wants to be super unicum that can carry 11 other players so having a below average winrate isn't entirely their fault?

 

haru stats.PNG

In the short run, no. A relatively small sample size could be luck had a greater impact than skill.

Once you're talking thousands of games, then luck should have little to do with it, but if you're looking at an individual ship, your sample size is still going be a fraction of your total number of games, and luck will still be a factor.

A better metric for figuring out how well or badly you're doing you have to keep track of yourself. Pay attention to your average finishing place on your team. Finish in the top half consistantly, and you're doing something right. Finish in the bottom half consistently, you need to look for places to improve.

  • Cool 1

Share this post


Link to post
Share on other sites
514
[KMS]
[KMS]
Beta Testers, In AlfaTesters
3,684 posts
12,254 battles
55 minutes ago, __Riptide_ said:

Title,

Here's an example, I have a garbage WR in my harugumo at 47%,, however at 81k average dmg per match, but does that mean that I suck in harugumo? I find it stupid that a player just judge another's stats by simply looking at the WR, They surely realize that not everyone are/ wants to be super unicum that can carry 11 other players so having a below average winrate isn't entirely their fault?

 

haru stats.PNG

Nope... its tier X which can flip so fast.

image.thumb.png.7a2ec18a495d361bccc9bfea110afa33.pngHere is my Harugumo stats.  Currently rerolling it again. 

You have almost past me in dmg in 19 games! You are hitting more accurately. I honestly have not played this boat like 2 times in 2 months as I was doing Research Bureau.  The Kit my best ship on that gunboat line at 57%. I also max its gun range.  Accuracy also depends on what you shoot. BB it should be higher.  Survialbity = More Exp.

 

Share this post


Link to post
Share on other sites
109
Beta Testers
431 posts
11,683 battles

Well...when I see 6 or 7 guys out of 12 on my team with 30% to %40wr /facepalm

" Statistics are very important for observation, analysis and mathematical prediction models. "

  • Confused 1

Share this post


Link to post
Share on other sites
575
[TIMT]
Members
1,134 posts
4,488 battles

From the bit of data I looked at in terms of how ships are doing and how players are doing, WR tells you part of the story but not everything.

For ships, there are several factors you could look at: avg. damage, avg. kills, K/D ratio, WR. For now, lets just focus on WR and average damage, for example lets look at Des Moines more specifically because she has been around almost forever and thus been played a lot. Keep in mind, the data is showing only people with 80+ matches.

desmoines_total.thumb.png.5b1c6555f2bfea661d09089170ee1e6b.png

80+ matches is really an arbitrary choice, though the idea is that below a certain number of games your numbers are not significant at all. In your case with 19 games and 47% WR this means you won 9 and lost 10. Imagine just a single game went a little different and you won that. Boom, up to 53% (10/19, rounded from 52.631...). So a single bit of good luck or bad luck skews the numbers considering we are looking for the range between 45-55% usually. Avg. damage is pretty well correlated with WR, especially at the lower end.

I would cautiously state that average damage is a bit more 'reliable' as it does not depend as much on other players (i.e. your team on the other flank not collapsing), however you might get better avg. damage in a loss since there is just way more targets for you to shoot at (quite paradoxical at first glance, I know).

Overall I would say the first 20-40 games are not really an indicator since they are subject to much luck, you getting used to a ship and possibly grinding modules etc. So after 50 games or so I would look at avg. damage and see where you are at, and after 100+ games WR might become accurate enough to be a good indicator, but not the measure. Overall, focus more on improving the stats like k/d or avg. damage you already have - no matter how good or bad -, since this will make you a better player and gives you a goal independent of winning/losing individual games. Helps with frustration, a lot.

Final words of warning: It is still a game, don't forget that. Take breaks, you need to have other games/hobbies etc. to not get too salty playing this game. Believe me, I have been there.

  • Cool 1

Share this post


Link to post
Share on other sites
1,160
[POP]
Members
4,749 posts
12,813 battles
3 minutes ago, Tekina_ said:

Over a large number of games if you are good in a ship you will generally win in it so solo win rate is a reasonable measure of skill. That's over a large number of games though. With small numbers of games it's hard for any statistic to really say whether you are good or not.

Definitely this.

40 minutes ago, __Riptide_ said:

Title,

Here's an example, I have a garbage WR in my harugumo at 47%,, however at 81k average dmg per match, but does that mean that I suck in harugumo? I find it stupid that a player just judge another's stats by simply looking at the WR, They surely realize that not everyone are/ wants to be super unicum that can carry 11 other players so having a below average winrate isn't entirely their fault?

 

haru stats.PNG

Sample size, a given skill in context with other stats, and even age of the stats are all important. With a small sample size (especially one as small as 19 battles), pretty much no stat is going to be a very accurate determination of skill. After 50 battles? 100 battles? 200 battles? Then your stats in a ship will be a better representation. But if you look at those stats in comparison to your overall stats in that ship type, a small sample size in one ship can still tell a lot - for example, you may only have 19 battles in Harugumo with a 47% WR, but if you have 2000+ battles in similar playstyle DDs and a low to mid 40% WR, I’d be more likely to say a lower stat count can still be indicative of performance (although that’s being based on a large sample size and context). But if you had good overall stats and lower ship stats with a small battle count, I’d look at those ship stats as still being influenced by luck and unfamiliarity with the ship.

Likewise, damage and xp can also influence how your stats reflect your performance. High WR but lower damage and average to good xp could point toward a DD player that plays well for the team (capping, spotting, anti DD work) but therefore doesn’t spend as much time farming. But if you’re a DD (especially something like the IJN gunboats) with a low WR and really high damage but average xp, that could mean you spend a lot of time farming damage off BBs and cruisers rather than helping your team/playing for the win. 
 

Age is also something that influences stats but often goes unmentioned in threads like this. For example, I have over 600 battles in NC (due to a period of more than a year early in my WoWS career where I only played that ship) with a 51% WR and average stats. However, the vast majority of those battles were from that year to year and a half when I was still relatively new to the game and before starting to really work to improve my skill and performance. In the past two years or so, in both divisions or solo randoms, I’ve had a significantly higher WR (at least in the upper 50s if not lower 60s on some days depending on teams and whether I personally have an off day or two) and better stats in that same ship. But if you look just at my overall NC stats, does that automatically mean I’m a bad player? Likewise my random stats for my entire account are sitting at a good-but-not-great almost 53%, but over the last few years (and especially since early to mid 2018), I’ve been pretty consistently pulling high 50s WRs in solo and anywhere from mid 60s to high 70s WRs in divs depending on streaks and who specifically I’m divisioning with. 
 

Bottom line, unless you’re looking to join one of the top competitive clans (in which case they’re going to be looking at overall stats as well as recent/current performance rather than just one or two low battle count ships), stats don’t matter. Personally, the main (and pretty much the only) reason I look at stats myself is to track my own performance and improvement. 

Plus, based on data shared by WG a year or two ago, the average randoms WR across the entire playerbase is (or was at that point) 48%, and I think it was something like 60-65% of the playerbase is in the 45-51/52% range. 

  • Cool 1

Share this post


Link to post
Share on other sites
560
[WKY04]
Members
591 posts
11,820 battles

Over time, yes, obviously. If you're a better player that will generally be reflected in WR. Too few battles can be misleading.

(Yes, I am completely average potato.)

Share this post


Link to post
Share on other sites
1,217
[FORM]
Members
1,969 posts
12,622 battles
1 hour ago, __Riptide_ said:

Title,

Here's an example, I have a garbage WR in my harugumo at 47%,, however at 81k average dmg per match, but does that mean that I suck in harugumo? I find it stupid that a player just judge another's stats by simply looking at the WR, They surely realize that not everyone are/ wants to be super unicum that can carry 11 other players so having a below average winrate isn't entirely their fault?

Not over the course of 19 battles in a ship, but after a hundred or so then yah I think it shows. Some ships fit a play style better, some are better at farming high damage but not contributing to the team as much, some perform worse against higher tiers than average, etc. But it is never entirely your fault, though over the course of a large enough sample trends definitely show the impact you have on our team.

I also believe in cursed ships that just seem to pull the worst teams. I blame it on the camo. 

Share this post


Link to post
Share on other sites
6,102
[WOLFG]
Members
31,377 posts
9,655 battles
1 hour ago, __Riptide_ said:

Title,

Here's an example, I have a garbage WR in my harugumo at 47%,, however at 81k average dmg per match, but does that mean that I suck in harugumo? I find it stupid that a player just judge another's stats by simply looking at the WR, They surely realize that not everyone are/ wants to be super unicum that can carry 11 other players so having a below average winrate isn't entirely their fault?

 

haru stats.PNG

You have 19 battles, it's too soon to tell. You could have been having a good streak right off the bat.

If you had a couple hundred games with those numbers, I would say you're good at dealing damage, but some of your teams may suffer by you playing CL instead of DD.

Share this post


Link to post
Share on other sites
243
[-CFI-]
Members
367 posts
3,786 battles
1 hour ago, __Riptide_ said:

Title,

Here's an example, I have a garbage WR in my harugumo at 47%,, however at 81k average dmg per match, but does that mean that I suck in harugumo? I find it stupid that a player just judge another's stats by simply looking at the WR, They surely realize that not everyone are/ wants to be super unicum that can carry 11 other players so having a below average winrate isn't entirely their fault?

The usefulness of statistics depends upon sample size. If the sample size is small, the stats are useless. If the sample size is large, the effects of confounding factors (such as the impact of other team members on wins and losses) cancel out. In other words, play 1000 games in a ship and you'll have close to 500 games with poor teams and 500 games with good teams. Then the stats tease out the typical strength of your impact on the game. But in order to draw even tentative conclusions, you need to play enough games that an outlier - either an unusually good or unusually bad performance by you, or a spectacularly bad or good team - won't skew the percentage to a significant amount.

Realistically, that really is something like 1000 games in a ship. After 100 you might be able to tentatively make judgements about whether your performance is improving or degrading, but the final number is still quite unreliable.

So basically, what I'm saying is that statistics don't matter... until they do. But at 19 games... your stats are meaningless. Almost every players' stats are meaningless. The only players whose stats are meaningful are that small subset of players who actually have thousands of games in individual ships, and then the stats for those players in those particular ships can say something interesting.

  • Cool 1

Share this post


Link to post
Share on other sites
560
[WKY04]
Members
591 posts
11,820 battles
4 minutes ago, TheArc said:

Not over the course of 19 battles in a ship, but after a hundred or so then yah I think it shows. Some ships fit a play style better, some are better at farming high damage but not contributing to the team as much, some perform worse against higher tiers than average, etc. But it is never entirely your fault, though over the course of a large enough sample trends definitely show the impact you have on our team.

I also believe in cursed ships that just seem to pull the worst teams. I blame it on the camo. 

Yes, if I use crazy camo like KOTS i inevitably catch a torpedo because I looked away or something.

Share this post


Link to post
Share on other sites

  • Recently Browsing   0 members

    No registered users viewing this page.

×